The two biggest names in machine control agree on more than their reputations suggest and disagree in exactly one region: where programs think. A Sinumerik 828D, Siemens’ compact-class control, and a Fanuc read the same motion spine identically, and the moment a program calls a repetitive operation or computes a value, the families part ways, with one numeric landmine buried at the border.
The shared spine, briefly
G0 through G3, plane selection, absolute and incremental modes, feeds and speeds, work offsets in the G54 family, compensation’s logic: the standard core reads the same on both, the convergence that makes machinists portable across the industry. An 828D program’s motion blocks need no translation for Fanuc-trained eyes, and that familiarity is precisely what makes the divergent layers surprising when they arrive.
Where the 828D itself sits matters for who meets this comparison: Siemens positions it as the compact-class Sinumerik, the control on the kind of standard mills and lathes that compete directly with Fanuc-controlled equivalents, which is why shops increasingly run both families on one floor and why operators cross between them weekly rather than once per career. The 840D’s deep-end territory, complex multi-axis work, has its own additional layers; the 828D comparison is the everyday one.
The three divergent layers
| Layer | Sinumerik 828D | Fanuc family |
|---|---|---|
| Repetitive operations | Named CYCLE calls: CYCLE81 drilling, CYCLE95 stock removal, parameters by name | G8x canned cycles, G7x lathe repeats, parameters by letter |
| Programmable logic | R-parameters (R0-R99…), GOTOF/GOTOB flow | Macro B: # variables, IF-GOTO, the protection layer |
| Unit selection | G70 inch / G71 metric | G20 inch / G21 metric |
| The same numbers elsewhere | , | G71 = lathe roughing cycle, G70 = finishing |
The cycle row is philosophy: a CYCLE81 call spells its parameters by name, self-documenting at the cost of length, where G81’s letter words are terse and universal. Neither is the primitive one; CAM posts to both transparently, and the difference lives at the control, in reading, editing, and hand-writing, where each family rewards its own fluency, the same split Siemens’ ShopMill conversational layer adds another floor to.
The logic row matters to anyone reading industrial programs: Siemens thinks in R-parameters with GOTOF/GOTOB flow, Fanuc in Macro B’s # variables, and the two systems’ programs do not translate mechanically, the logic gets rebuilt per family.
The landmine, in full
G70 and G71 are the crossover trap because both families use both numbers for unrelated jobs: on Sinumerik they select inch and metric input, on Fanuc lathes they are the finishing and roughing cycles. A Fanuc-trained reader meeting G71 atop a Siemens program and parsing it as a roughing call has misread the program’s units, the most basic fact it states; the reverse migration misreads a roughing cycle as a unit switch. No mnemonic fixes an overloaded number; the protection is knowing the overload exists, the same number-reuse awareness the lathe codes keep teaching, and reading every cross-family program with the dialect question already open.
Crossing between the families
The preparation is asymmetric in the right way: the shared core transfers untouched, kept automatic by the free 60-second rounds on the G-code practice page, and the per-family layer is one documentation read plus one working program studied. For the Fanuc-trained machinist arriving at an 828D: recalibrate G70/G71 first, learn to read CYCLE calls (they are friendlier than they look, the names say what the letters used to hide), and treat R-parameter logic as a new dialect of familiar thinking. For the Siemens-trained machinist headed the other way, the same list mirrored, with Macro B’s terseness as the adjustment. Both directions confirm the industry’s quiet rule: controls compete at the edges and cooperate at the core, and the core is where careers live.
Sources
Frequently asked questions
What are the main G-code differences between Siemens 828D and Fanuc?
The motion spine is shared; the divergence concentrates in three layers: named CYCLE calls versus G8x/G7x cycles, R-parameters with GOTO versus Macro B, and conventions led by the G70/G71 landmine.
What is the G71 trap between Siemens and Fanuc?
The same number does unrelated jobs: inch/metric selection on Sinumerik, the lathe roughing cycle on Fanuc. Misreading it misreads a program’s most basic facts, so it is the first recalibration when crossing families.
Are Siemens CYCLE calls better or worse than Fanuc canned cycles?
Different philosophy, same job: named parameters read self-documenting, letter words run terse and universal. CAM posts to both transparently; the difference lives at the control.
How should a Fanuc-trained machinist prepare for an 828D shop?
Keep the core automatic, the free G-Code Sprint app’s 60-second rounds handle that, then recalibrate G70/G71, learn CYCLE-call reading, and treat R-parameter logic as a new dialect of familiar thinking.